International Journal of Agricultural Science and Technology (IJAST) Volume 1 Issue 4, November 2013 www.seipub.org/ijast

Isolation, Sequence Identification and Tissue Expression of a Novel Tobacco (Nicotiana tabacum) Gene-TBG1 Nian Fuzhao1, Zhang Yonghui2, Su Xiangyun3, Zou Juqiu3, Zhao Leifeng*1

1College of Tobacco Science, Yunnan Agricultural University, Kunming 650201, China 2Technological Center of Tobacco Company of Sichuan Province, Luzhou 646000, China 3Feedstock Supply Department of Industrial Limited-Liability Company of Chinese Tobacco Company of Hunan province, Changsha 410014, China *[email protected]

Abstract 1994; Carey et al. 1995; Smith et al. 1998; Smith and The complete coding sequence of one tobacco (nicotiana Gross, 2000; Carey et al. 2001; Shlush et al. 2013; tabacum) gene-β-galactosidase precursor (TBG1), was Fonseca-Berzal et al. 2013; Choi et al. 2013). amplified based on some tobacco ESTs. The complete coding Although β-galactosidases play important roles in sequence of tobacco TBG1 gene was 2,526 bp which encodes biological processes of plants, until today, the tobacco a protein of 841 amino acids. Sequence analysis revealed that β-galactosidase precursor (TBG1) gene has not been the TBG1 of tobacco shares high homology with the TBG1 of reported yet. In present experiment, we would isolate lycopersicon esculentum (91%), cacao (81%), castor bean TBG1 (81%) and papaya (81%). Results also showed that tobacco the complete mRNA sequence of tobacco gene, TBG1 gene has a closer genetic relationship with the TBG1 subsequently perform some necessary sequence gene of lycopersicon esculentum. The expression profile was analysis and tissue expression analysis for this gene; studied and the results indicated that tobacco TBG1 gene was thus the primary foundation of understanding this differentially expressed in detected tobacco tissues including tobacco gene was established. leaf, stem, root and flower. Our experiment established the foundation for further research on this tobacco gene. Materials and Methods

Keywords Samples Collection, RNA Extraction and First-strand Tobacco; TBG1; Tissue Expression cDNA Synthesis Tobacco plants (Chinese local variety Yunyan 85) were Introduction grown in a naturally lit glasshouse with normal β-galactosidases are thought to be responsible for the irrigation and fertilization. The tissues including leave, conspicuous reduction in the level of galactosyl residues stem, root, flower were harvested and immediately which occurs in many ripening fruits. These enzymes frozen in liquid nitrogen and stored at -80℃. Total have also been detected in a wide range of plant RNA extraction and first-strand cDNA synthesis for organs and tissues undergoing developmental changes these tissue samples were performed as the methods such as seeds, cotyledons, elongating epicotyls, and in described by Li et al. (2008). freshly harvested asparagus spears. So far, only exo- Isolation of the Coding Sequence acting activities have been found in plants and these enzymes are characterized by their ability to hydrolyse RT-PCR was performed to amplify complete coding terminal non-reducing β-d-galactosyl residues from sequences of tobacco TBG1 gene using the cDNA β-d-galactosides. They are usually assayed by their obtained from the pooled tissues above. The 20 µl ability to release galactosyl residues from artificial reaction system was: 2.0 µl cDNA, 2.0 µl 2 mM mixed substrates such as p-nitrophenyl-β-d-galactopyranoside, dNTPs, 2.0 µl 10×Taq DNA polymerase buffer, 1.2 µl 4-methyl umbelliferyl-β-d-galactopyranoside and X-gal. 25 mM MgCl2, 1.0 µl 10 mM forward primer, 1.0 µl 10 In many cases, however, their natural substrates are mM reverse primer, 2.0 units of Taq DNA polymerase unknown, but probably include chloroplast galactolipids (1 U/1µl), and 9.8 µl sterile water. The primers for as well as cell wall galactans(Gross and Sams,1984; MS , tobacco TBG1 gene isolation were designed based on

57 www.seipub.org/ijast International Journal of Agricultural Science and Technology (IJAST) Volume 1 Issue 4, November 2013 the tobacco EST sequences (GeneBank numbers: Quantitative Real Time PCR (qRT-PCR) for Tissue FS390322, FG200071, FG192756, FG191482, FG164696 Expression Profile Analysis and EB678821) which were highly homologues with qRT-PCR for evaluating the level of mRNA for TBG1 the coding sequence of TBG1 gene of lycopersicon gene was performed on the ABI Prism 7300 Sequence esculentum(TABLE 1). The PCR program initially Detection Systems (Applied Biosystems, Foster City, started with a 94°C denaturation for 4 min, followed CA, USA). PCR reactions for each sample were carried by 35 cycles of 94°C/1 min, 56°C (TABLE 1)/1 min, out in 25 µl reaction volume containing 1µl SYBR 72°C /1 min, then 72°C extension for 10 min, finally 4°C to terminate the reaction. Every PCR was repeated Green real-time PCR Master Mix, 100 ng cDNA five times. The PCR product was then cloned into template and 200 nM each primer. Conditions for PMD18-T vector and sequenced bidirectionally with real-time PCR were: an initial denaturation at 95°C for the commercial fluorometric method. At least five 3 min, 40 cycles of 95°C for 15 s, optimal annealing independent clones were sequenced. temperature for each specific primer for 15 s (TABLE 2), 72°C for 20 s. For each sample, reactions were set up in TABLE 1 RT-PCR PRIMERS FOR TOBACCO TBG1 GENE AND ANNEALING TEMPERATURE triplicate to ensure the reproducibility of the results. The gene relative expression levels were quantified Gene Primer sequence Ta/ °C Length/(bp) Forward 1: 5’- relative to the expression of the reference gene, actin ATGGATTGTTCTTGGATTGC-3’ (GenBank Accession No. GQ339768), by employing the TBG1 56 2526 Reverse 1: 5’- 2 -ΔΔCt value model (Livak and Schmittgen TD, 2001). TCAGCTGCAAACAACCTC-3’

TABLE 2 QRT-PCR PRIMERS FOR TOBACCO TBG1, ACTIN GENES AND Sequence Analysis ANNEALING TEMPERATURE The cDNA sequence prediction was conducted using Gene Primer sequence Ta/ °C Length/(bp) GenScan software (http://genes.mit.edu/ GENSCAN .html). Forward 2: 5’- AGAAGGAGGAGTGGATGT-3’ The protein prediction and analysis were performed TBG1 56 224 Reverse 2: 5’- using the Conserved Domain Architecture Retrieval TGAAACTGATACCTGGAAC-3’ Tool of BLAST at the National Center for Forward 3:5’- Biotechnology Information (NCBI) server (http://www. CCATTCTTCGTTTGGACCTT -3’ Actin 56 257 Reverse 3: 5’- ncbi.nlm.nih.gov/BLAST) and the Clustalw software TTCTGGGCAACGGAACCT-3’ (http://www.ebi.ac.uk/clustalw).

58 International Journal of Agricultural Science and Technology (IJAST) Volume 1 Issue 4, November 2013 www.seipub.org/ijast

FIG. 2 THE ALIGNMENT OF THE PROTEINS ENCODED BY TOBACCO TBG1 GENE AND FOUR OTHER KINDS OF TBG1 PROTEINS .

59 www.seipub.org/ijast International Journal of Agricultural Science and Technology (IJAST) Volume 1 Issue 4, November 2013

Based on the results of the alignment of different Results and Discussion species of TBG1 proteins, a phylogenetic tree was Isolation Result for Tobacco TBG1 Gene constructed using the ClustalW software (http://www. ebi.ac.uk/clustalw), as shown in FIG. 3. The For tobacco TBG1 gene, through RT-PCR with pooled phylogenetic tree analysis revealed that the tobacco tissue cDNAs, the resulting PCR products were 2,526 TBG1 gene has a closer genetic relationship with that bp (FIG. 1). of lycopersicon esculentum.

Tissue Expression Profile Tissue expression profile analysis was carried out and results revealed that the tobacco TBG1 gene was moderately expressed in flower and leaf, but hardly expressed in root and stem (FIG. 4).

FIG. 1 PCR RESULT FOR TOBACCO TBG1 GENE. M DL2000 DNA MARKERS; 1, PCR PRODUCT FOR TOBACCO TBG1 GENE

Sequence Analysis These cDNA nucleotide sequences analysis using the BLAST software at NCBI server (http://www.ncbi. nlm.nih.gov/BLAST) revealed that this gene was not homologous to any of the known tobacco gene and it was then deposited into the Genbank database (Accession number: KF267437). FIG. 4 EXPRESSION ANALYSIS OF TBG1 GENE MRNA IN The sequence prediction was carried out using the VARIOUS TISSUES GenScan software and results showed that the 2,526-bp Comparative genomics research has revealed that cDNA sequence represents one single gene which virtually all (99%) of the protein-coding genes in encodes 841 amino acids. The theoretical isoelectric humans align with homologs in mouse, and over 80% point (pI) and molecular weight (Mw) of the deduced are clear 1:1 orthologs for human and mouse both proteins of this tobacco gene were also computed belong to mammalian (Hardison, 2003). This extensive using the Compute pI/Mw Tool (http://www.expasy. conservation in protein-coding regions implied that org/tools/pi_tool.html). The pI of tobacco TBG1 is 6.68. this conservation of protein-coding sequences may be The molecular weight of this putative protein is expected in tobacco and lycopersicon esculentum as 93674.67. they are both plants of solanaceae. From sequence Further BLAST analysis of these proteins revealed that analysis of TBG1 genes, it can be seen that the coding tobacco TBG1 has high homology with the TBG1 of sequences of TBG1 genes were highly conserved in lycopersicon esculentum (Accession number: NP_ some plants. This implied that the some plants can be 001234303, 91%), cacao (Accession number: EOY27097, utilized as model organisms and some tobacco genes 81%), castor bean (Accession number: XP_002527409, can be isolated based on the coding sequence 81%) and papaya (Accesession number: ACP18875, information of these plants. 81%) (FIG. 2). The phylogenetic tree analysis revealed that the tobacco TBG1 gene has a closer genetic relationship with that of lycopersicon esculentum. This implied that lycopersicon esculentum can be regarded as model organism to study the tobacco TBG1 gene or tobacco can be utilized as model organism to study the lycopersicon esculentum TBG1 gene. From the tissue distribution analysis in our experiment, it can be seen that TBG1 gene was obviously

differentially expressed in some tissues. The suitable FIG. 3 THE PHYLOGENETIC TREE FOR FIVE KINDS OF TBG1 GENES explanation for this under current conditions is that at

60 International Journal of Agricultural Science and Technology (IJAST) Volume 1 Issue 4, November 2013 www.seipub.org/ijast

the same time those biological activities related to the (2013): 4851-4856 mRNA expression of this gene were presented Gross KC., and Sams CE. “Changes in cell wall neutral sugar diversely in different tissues. composition during fruit ripening: a species survey.” In conclusion, the tobacco TBG1 gene was firstly Phytochemistry 23(1984):2457–2461 isolated and then tissue expression profile analysis was Hardison RC. “Comparative genomics.” PLoS Biol. 1 (2003): performed; thus the primary foundation was established for further research on this tobacco gene. E58. Li Y., Meng F., Yin J., Liu H., Si Z., Ni Z., Sun Q., Ren J., Niu REFERENCES H. “Isolation and comparative expression analysis of six MBD genes in wheat.” Biochim Biophys Acta1779 Carey AT., Holt K., Picard S., Wilde R., Tucker GA., Bird CR., (2008):90–98 Schuch W., Seymour GB. “Tomato exo-(1-4)-β-d- Livak KJ., and Schmittgen TD. “Analysis of relative gene galactanase: isolation, changes during ripening of normal expression data using real-time quantitative PCR and the and mutant tomato fruit and characterization of a related 2(-Delta Delta C(T)) Method.” Methods 25(2001): 402-408. cDNA clone.” Plant Physiology108(1995):1099–1107 MS, Reid JSG. “Purification and properties of a novel Carey AT., Smith DL., Harrison E., Bird CR., Gross KC., β-galactosidase or exo‐(1‐4)-β-d-galactanase from the Seymour GB., Tucker GA. “Down-regulation of a cotyledons of germinated Lupinus angustifolius L. ripening-related beta-galactosidase gene (TBG1) in seeds.” Planta 192(1994): 502–511 transgenic tomato fruits.” J Exp Bot.52(2001):663-668. Shlush LI., and Selig S. “Digital Image Analysis of Cells Choi O., Lee Y., Han I., Kim H., Goo E., Kim J., Hwang I. “A Stained with the Senescence-Associated β-Galactosidase simple and sensitive biosensor strain for detecting Assay.” Methods Mol Biol 1048(2013):11-18. toxoflavin using β-galactosidase activity.” Biosens Smith DL., and Gross KC. “A family of at least seven Bioelectron 50C(2013): 256-261 β-galactosidase genes is expressed during tomato fruit Fonseca-Berzal C., Merchán Arenas DR., Romero Bohórquez development.” Plant Physiology123(2000):1173–1184 AR., Escario JA., Kouznetsov VV., Gómez-Barrio A. Smith DL., Starrett DA., Gross KC. “A gene coding for “Selective activity of 2,4-diaryl-1,2,3,4-tetrahydroquinolines on Trypanosoma cruzi epimastigotes and amastigotes tomato fruit β-galactosidase II is expressed during fruit ripening.” Plant Physiology 117(1998):417–423 expressing β-galactosidase.” Bioorg Med Chem Lett. 23

61 www.seipub.org/ijast International Journal of Agricultural Science and Technology (IJAST) Volume 1 Issue 4, November 2013

Characterization of Regolith Salt Storage in a Small Catchment of the Basin Nebo Jovanovic*1, Richard D.H. Bugan1, Ashton Maherry1, Louise Soltau1, Willem P. De Clercq2, Martin V. Fey2

1Natural Resources and Environment, CSIR, PO Box 320, Stellenbosch 7599, 2Department of Soil Science, University of Stellenbosch, Private Bag X1, Matieland 7602, South Africa *[email protected]

Abstract occurrence of dryland salinity is widespread Dryland salinity recharge from agricultural land has been throughout semi-arid regions of the world, with identified as the main source of salinization in the Berg river associated impacts on water resources, ecosystems and basin (, South Africa), which represents a land productivity (Pannell, McFarlane and Ferdowsian, challenge to water users and water quality managers. The 2001; Clarke, George, Bell and Hatton, 2002). The aim of this study was to characterize and quantify regolith process of dryland salinization occurs either through salt storage in a small catchment (~ 20 ha) representative of weathering of rock minerals or salts may be brought the saline environment and drier bio-climatic conditions of into the landscape from the ocean, by rain or wind the mid- and lower reaches of the Berg river basin, where storage and potential discharge of salts are likely to be the (Hingston and Gailitis, 1976; Acworth and Jankowski, greatest. Electromagnetic induction measurements with an 2001). Studies conducted in Australia have produced EM-38 meter indicated that salinity in the landscape is evidence that dryland salinity may be exacerbated by driven by the topo-sequence, anti-erosion contours act as human activities (Greiner, 1998; Stirzaker, Cook and barriers to water and salt fluxes, highly saline scalds are Knight, 1999; Angus, Gault, Peoples, Stapper and Van associated with convex hillslopes, and soil salinity is lower Herwaarden, 2001; Clarke, George, Bell and Hatton, at the end of the rainy winter season compared to that at the 2002). In particular, clearing of natural perennial dry summer. Resistivity tomography measurements taken scrubland to make way for cultivated crops and with an ABEM Terrameter SAS1000 ground imaging apparatus down to approximately 70 m depth indicated that pastures has resulted in changes in the water balance, subsurface drainage features are often associated with rising groundwater tables and mobilization of fossil salinity and topographic flow pathways, as a result of shale connate salts (Pannell and Ewing, 2006). weathering. Analyses of core samples collected during borehole drilling indicated that the regolith contains about In South Africa, dryland salinity is common in the 400 tons of salts per hectare, stored mainly as a bulge in the Berg river catchment (Western Cape Province). The unsaturated vadose zone (between about 5 and 10 m depth), South African Department of Water Affairs (formerly above the water table and below the soil zone, discharging Department of Water Affairs and Forestry) has mainly through lateral leaching by groundwater from the monitored Berg river salinization since the mid 1970’s. phreatic zone. Groundwater quality becomes increasingly A cycle of research projects funded by the Water saline from the recharge area (0.68 dS m-1) towards the Research Commission (Fey and De Clercq, 2004; bottom end of the catchment (3.75 dS m-1), and it is a mixed Gorgens and De Clercq, 2005; De Clercq, Jovanovic cation-chloride-type. The effects of different climates, hydrogeological characteristics and land uses on the salt and Fey, 2010) indicated that irrigated agriculture does discharge into the mid- and lower reaches of the Berg river not account for the Berg river salinization, rather need to be investigated. salinity recharge from dryland agriculture is the main source of salinity. This is an enormous challenge Keywords because of the variety of water users (drinking water Dryland Salinity; Electromagnetic Induction; Groundwater supply, agricultural, industrial, ecosystems, recreational), Quality; Regolith; Resistivity Tomography each requiring specific volumes and qualities of water (DWAF, 2004). Characterization and quantification of Introduction natural salts in the regolith is therefore important in Salinization of soil and water resources is usually order to determine the extent and hazard of associated with irrigated agriculture (Van Rensburg, salinization, to identify saline scalds as major De Clercq, Barnard and Du Preez, 2011). However, the contributors to runoff salinity (Hughes, Khan, Crosbie,

62 International Journal of Agricultural Science and Technology (IJAST) Volume 1 Issue 4, November 2013 www.seipub.org/ijast

Helliwell and Michalk, 2007), as well as to model and small catchment is located in quaternary catchment predict salinity processes. G10F, adjacent to and with a similar hydrogeological set-up as Sandspruit. Estimates of salt accretion and storage in the unsaturated regolith and groundwater in Western Site Description Australia were made by Johnson (1987) and McArthur, Chrurchward and Hick (1977), whilst Engel, McFarlane and Street (1989) used geophysics and drilling to characterize flow pathways (in Clarke, George, Bell and Hatton, 2002). Acworth and Jankowski (2001) carried out perhaps the most comprehensive study on the source of salts causing dryland salinity in a small catchment in New South Wales (Australia) by using drilling, geophysics, hydrogeochemistry and groundwater monitoring. Acworth (1999) used electromagnetic ground conductivity meters to obtain 2-dimensional vertical sections of salinity-induced apparent electrical conductivities. These data could then be correlated to electrical conductivity (EC) values obtained from soil solution and borehole drilling core extracts, groundwater quality and other field data. Bennett and George (1995) measured soil salinity with an electromagnetic induction meter in order to determine the salinity effects on growth of Eucalyptus plantations planted to lower groundwater tables in South-Western Australia. Stirzaker, Cook and Knight (1999) supplied analytical expressions to determine salt accumulation in the root zone from shallow water tables under different agroforestry systems. Other Australian studies were focused on relating the increase in water table level to the extent of land salinization. For example, Greiner (1998) used a spatial optimization model for analyzing catchment management (SMAC) to predict salinization as affected by current land use, optimized land use scenario, and including climatic variability. Clarke, George, Bell and Hatton (2002) discussed extensively how the geology and land use changes in South- Western Australia affect water flow pathways and recharge. In the Berg river catchment, salt storage was seldom measured. The only comprehensive study done so far was that of Flugel (1995), who used drilling, soil mapping as well as measurements of soil water and groundwater salinity to quantify the storage of salts at several locations and estimated the salt balance in the FIG. 1 (a) LOCATION OF THE STUDY AREA IN AFRICA, (b) Sandspruit catchment. Sandspruit is a tributary of the GEOLOGICAL MAP OF THE BERG RIVER BASIN (1:250,000, AVAILABLE AT http://www.geoscience.org.za, ACCESSED 26 2 Berg river approximately 150 km in size, located in MARCH 2009) AND (c) LOCATION OF THE GOEDERTROU quaternary catchment G10J as delineated by the SMALL CATCHMENT IN Google Earth. THE POSITIONS OF Department of Water Affairs. The aim of the current BOREHOLES (BH), RESISTIVITY MEASUREMENT TRANSECTS study was to characterize and quantify the salts stored (DASHED LINES RT), ELECTROMAGNETIC INDUCTION MEASUREMENT TRANSECT (SOLID LINE EM-38), STOCK- in the regolith of a small catchment representative of WATERING DAM AND MAN-MADE CONTOURS ARE the saline environment in the Berg river basin. The INDICATED

63 www.seipub.org/ijast International Journal of Agricultural Science and Technology (IJAST) Volume 1 Issue 4, November 2013

The source of the Berg river is in the Franschhoek and storage and potential discharge of salts are likely to be Jonkershoek mountains (East of ) and it the greatest. The Goedertrou catchment receives less flows in a north-westerly direction where it eventually than the average rainfall for the Berg river catchment discharges into the sea at Velddrif (Fig. 1b). The Berg due to topographic conditions (approximately 330 mm river is approximately 270 km in length and has a a-1) with an average annual potential evaporation of catchment size of about 900 km2 (DWAF, 1993). A 1615 mm. The general slope is from the South-West to large part of the catchment is located in semi-arid the North-East with the main surface drainage wheat lands which receive most of their rain during features described by the man-made anti-erosion the winter months (450 mm a-1 on average, mainly contours visible in Fig. 1c. from May to October). The geology of the Berg river Pedological, chemical and physical characteristics of basin is dominated by the Malmesbury Group shales the soils were characterized in approximately 1.2 m (western banks, grey shade in Fig. 1b) and the Table deep pits, excavated on a 50x50 m grid. The dominant Mountain Group sandstones (eastern banks, light soil forms are Glenrosa (Soil Classification Working brown shade in Fig. 1b). Group, 1991)-Cambisol (FAO, 1998) and Swartland The Malmesbury Group is a proterozoic marine (Soil Classification Working Group, 1991) - Luvisol deposit comprising greywacke and phyllite beds, with (FAO, 1998). The detailed pedological, chemical and beds and lenses of quartz schist, limestone and grit. physical information on the soils at Goedertrou were The Malmesbury Group Aquifer occurring in the area reported by De Clercq, Jovanovic and Fey (2010). is classified as a Minor Aquifer System (Parsons, 1995). This is defined as fractured or potentially fractured Material and Methods rocks that do not have a high primary permeability, or The approach used to characterize salinity and salt other formations of variable permeability. Their extent storage in the Goedertrou small catchment included may be limited and water quality variable. Although electromagnetic induction measurements, geophysics these aquifers seldom produce large quantities of and borehole establishment. water, they are important for both local supplies and discharge as baseflow into rivers. They also have a Electromagnetic Induction moderate vulnerability to pollution. Regional Electromagnetic induction is a non-invasive technique groundwater flow is towards the Berg river, and that measures electrical conductivity by inducing an thereafter, parallel to the river in the western part of electrical field in the soil. Apparent electrical the catchment (Fig. 1b). conductivity of the bulk soil is measured and this In order to characterize and quantify the regolith salt represents a combination of soluble salts, clay content storage in the typically saline environment, a small and mineralogy, porosity, structure, bulk density, catchment located on the Goedertrou farm was cation exchange capacity, soil water content and selected, 5 km North-East of the town of temperature (McNeill, 1980; Corwin and Plant, 2005). and approximately 70 km North of Cape Town in the The EM-38 electromagnetic induction meter (Geonics Western Cape Province. The Google Earth view of the Ltd., Missisauga, Ontario, Canada) was used to map Goedertrou catchment is shown in Fig. 1c (latitude soil salinity in the Goedertrou catchment. The 33°18’33.16” S; longitude 18°53’40.67” E; altitude 120 measurements were taken during April 2005, on a 1 m m at a stock-watering dam collecting runoff and x1 m grid. The meter was positioned vertically in interflow water). The Goedertrou catchment is ~20 ha order to scan a soil depth of between 0.15 m and 1.5 m. in size and deemed to be representative of the area in A transect on the North-facing slope of the catchment terms of soils (residual soils with differing sand, clay was then scanned in October 2006 (end of rainy winter and gravel contents generally overlie silcrete and season) and April 2007 (end of dry summer season) at weathered Malmesbury Group shales), topography approximately 10 cm intervals. The meter was (slope ~10% on average), land use (dryland winter positioned vertically and horizontally in order to scan wheat cropping followed by two years of fallow two soil depths, namely 0.15-1.5 m (vertical reading) grazing; and vineyard) and land management (man- and 0.3 m (horizontal reading). The location of the made anti-erosion contours; shallow tillage to break transect is shown in Fig. 1c. The purpose of the up the soil surface crust). In addition, the site is transect measurements was to assess the seasonal representative of the drier bio-climatic conditions in change in bulk soil salinity, as well as the link between the mid- and lower reaches of the Berg river, where salt distribution in the landscape, salinity profile and

64 International Journal of Agricultural Science and Technology (IJAST) Volume 1 Issue 4, November 2013 www.seipub.org/ijast topography. parallel to topographic contours were therefore used to identify geologically-driven drainage patterns. Resistivity Tomography Borehole Drilling and Sampling Resistivity tomography is a non-invasive technique that was used to provide a pseudo-section of change in Following the results obtained from resistivity and the electrical properties in the subsurface along a specified identification of geologically-driven drainage patterns, line (transect). The bulk resistivity of porous rocks nine boreholes were installed in the Goedertrou varies mostly because of changes in porosity and catchment from March to April 2005 in order to obtain saturation of the host rock and salinity of the pore background information on the aquifers. The fluid (Telford, Geldart, and Sheriff, 1990), in the boreholes were drilled using a rotary percussion rig. absence of clay. In the presence of clay (from Undisturbed and disturbed unsaturated zone samples weathering of the Malmesbury shale in this case), a were taken during drilling down to the water table or surface conduction term has to be included (Bussian, to a maximum depth of 25 m. The samples were taken 1983). Relatively, the high clay content and high at either 0.5 m (tube method) or 1 m (auger-tip method) salinity of the pore fluid will result in a very low interval depths. Fifty tube and 80 auger-tip samples resistivity for the subsurface. were taken in total. All samples were sealed The resistivity measurements were taken with an immediately after being taken in order to prevent ABEM Terrameter SAS1000 ground imaging contamination and evaporation. A full set of apparatus. The measurement sequence is controlled by laboratory chemical analyses was done on the the ABEM electrode selector ES464 (ABEM Instrument saturated paste extracts of these regolith core samples. AB, Sweden). Two multi-core cables with 21 electrode Borehole information was logged for all boreholes. take-outs every 10 m were used. The cables were laid Borehole casing was 125 mm diameter Class 12 PVC out on the ground, consecutively from start to end, in (with stabilisers) to termination depth, with screened a straight line to cover a transect of 400 m. An sections placed adjacent to water strikes. Sand was electrode (metal stake) was inserted into the ground used as stabilization material adjacent to screened next to every electrode take-out on the cable. The intervals, whilst gravel was positioned on solid electrode take-out was then connected to the electrode sections. Bentonite seals were also installed where with a short cable jumper. The data were collected appropriate in order to prevent shallow groundwater using a standard protocol of the Schlumberger- from entering into boreholes as well as the ingress of Wenner array (Loke, 2001). All data were acquired for surface runoff. Groundwater was sampled after a = 1 to 8 and n = 1, 2 and 4, where ‘a’ is the electrode borehole installation was completed. As part of this separation multiplication factor and ‘n’ is the electrode process, boreholes were purged with a submersible spacing multiplication factor. The apparent resistivity pump at a rate of 0.75 L s-1 for a predetermined data acquired in the field were inverted using the amount of time, equal to three times the borehole RES2DINV software (Loke, 2001) to provide a pseudo- casing volume. The groundwater samples were then section of changes in resistivity along each profile. taken, placed immediately into a cooler box in the field Resistivity measurements were taken during and transferred to a refrigerator in the laboratory on September 2005, along the transects shown in Fig. 1c. the same day of sampling. A full set of laboratory The depth of investigation for this particular set-up chemical analyses was done on the groundwater was about 70 m. The resistivity method was applied to samples. define subsurface drainage. The envisaged drainage pattern would be reflected in deeper channels of Results weathering in the underlying shale which is related to Electromagentic Induction fracture zones. It was expected that surface drainage in the small catchment is partly controlled by A soil salinity map for the Goedertrou catchment was geological characteristics. The transect locations were developed using measurements taken with the EM-38 selected in such a way to cross expected drainage meter in April 2005 (Fig. 2). The data collected on a 1 patterns perpendicularly, as well as to minimize the m x1 m grid represented the vertical scan (0.15-1.5 m effects of increased wetness (caused by damming of depth), and these were interpolated using ordinary anti-erosion contours) on the resistivity readings. The kriging. The map shows that bulk soil salinity varied resistivity measurements in transects generally between approximately 20 and 160 mS m-1 across the

65 www.seipub.org/ijast International Journal of Agricultural Science and Technology (IJAST) Volume 1 Issue 4, November 2013 small catchment (Fig. 2). The Goedertrou catchment represent the average bulk soil EC at each measuring has a soil mantle that is variably saline and this is point over the measuring depth. The square blocks in possibly attributable to salt influx from an extensive Fig. 3 indicate the position of the contour banks. recharge zone at the top of the catchment or to In general, bulk soil salinity values were lower in exposure of saline saprolite through natural erosion. October 2006 compared to April 2007. The saline scald The poorly structured, low carbon topsoil (with a clay identified in the salinity map (Fig. 2) is also fraction consisting chiefly of kaolinite and mica), recognizable in the mid-slope area of the transect (Fig. although lighter textured than the subsoil, exhibits 3; 150-200 m distance from base of valley). It is also strong crusting. Runoff during rain events is clear that salinity in the landscape varies at different consequently intense (De Clercq, Jovanovic and Fey, sections of the topo-sequence and that anti-erosion 2010). The more clayey subsoil, low hydraulic contours have a marked effect on salinity distribution conductivity and high compaction (average bulk in the catchment. density in the B-horizon, where present, is 1.89 g cm-3) represents a further barrier to groundwater recharge, 120 3.5 and lateral migration of perched water through the 100 3 surface horizon is therefore common. 2.5 80 2 60 1.5 40 v/h ratio 1

Bulk soil electrical 20

conductivity (mS/m)conductivity 0.5 0 0 0 50 100 150 200 250 300 Distance from base of valley (m) Horizontal Vertical Contours v/h ratio

(a)

120 3

100 2.5

80 2

60 1.5

40 1 v/h ratio

Bulk soil electrical 20 0.5 conductivity (mS/m)conductivity FIG. 2 EM-38 SCAN OF THE GOEDERTROU CATCHMENT 0 0 -1 INDICATING SALINITY (mS m ) AND THE SPATIAL 0 50 100 150 200 250 300 DISTRIBUTION THEREOF IN THE CATCHMENT (AFTER DE Distance from base of valley (m) CLERCQ, JOVANOVIC AND FEY, 2010) Horizontal Vertical Contours v/h ratio

Bulk soil salinity was generally higher on the North- (b) facing slope, in particular at a convex position on the FIG. 3 BULK SOIL ELECTRICAL CONDUCTIVITY VERTICAL mid-slope, where salinity effects were visible in the AND HORIZONTAL READINGS TAKEN WITH AN EM-38 form of a saline scald where wheat failed to germinate. METER ON A NORTH-FACING HILLSLOPE TRANSECT BEFORE In most of these very saline spots, the B horizon is (APRIL 2007 (a)) AND AFTER WINTER RAINS (OCTOBER 2006 absent leaving a very thin A horizon directly on shale. (b)), AND SMOOTHED RATIOS OF VERTICAL AND HORIZONTAL (v/h) READINGS (MOVING POINT AVERAGE OF Down-slope leaching from the crest and down-valley NINE READINGS) (AFTER CLERCQ, JOVANOVIC AND FEY, 2010) leaching from the foot-slope soils could possibly The smoothed ratio of vertical and horizontal readings account for this mid-slope salinity maximum. (moving point average of nine readings) were also The vertical and horizontal scans obtained with the compared for October 2006 and April 2007 in order to EM-38 meter on the North-facing hillslope transect investigate differences in soil salinity profiles along (Fig. 1c) are shown in Fig. 3. The measurements were the hillslope transect (Fig. 3). The ratio declined and taken in late winter (October 2006) when most salts approached 1 towards the base of the valley meaning were washed away and/or back into the soil through that, lower down in the catchment, a homogeneous infiltration, and at the end of summer (April 2007) profile of salt seemed to be the norm. In the upper part when the soils were driest and a lot of salt occurred on of the catchment, the vertical reading was much and/or close to the soil surface. The EM-38 data higher than the horizontal reading, indicating much

66 International Journal of Agricultural Science and Technology (IJAST) Volume 1 Issue 4, November 2013 www.seipub.org/ijast more salts lower down in the soil profile compared to (weathered shale) and fractured shale has been the soil surface layer. interpolated from the borehole log information for BH 1. The northern half of the profile shows a lower Resistivity Tomography resistivity at intermediate depth which is likely to The inverted resistivity profiles for transects RT1 to correspond to high salinity pore fluid. Along this RT4 (Fig. 1c) are shown in Fig. 4. The resistivity scale profile, there are 2 likely subsurface fracture zones along all profiles has been normalised to aid (features 1 and 2) related to subsurface drainage. It is comparisons. The distance and depth along all profiles not clear whether these correspond to surface drainage are approximately the same. In general, all the profiles in any way. The underlying higher resistivity layer show a shallow very low to low resistivity layer (1 to corresponds to the fractured shale that grades into ~250 ohm.m) that corresponds to the weathered unweathered shale. The resistivity cross-section also Malmesbury Shale (clay). This is underlain by a higher shows two high resistivity sections along the shallow resistivity layer (> 250 ohm.m) corresponding to the part of the profile. This does not occur on any of the fractured shale that grades into unweathered shale at other profiles. The northern part of the profile is greater depths. This correlates well with previously started in a silcrete lens that is part of the paleo- established resistivity values for shale; and from weathering African Surface. This accounts for the high previous resistivity studies in shale, it has been resistivity at shallow depth, however the change in established that the fractured shale has a resistivity of resistivity towards the end of the profile is not very between 200 and 500 ohm.m (Soltau and Anderson, clear. This could be an indication of transported 2006). The resistivity for unweathered shale is often silicate material, which has a much higher resistivity above 500 ohm.m, but usually does not exceed 2000 than the clay material. It could also in itself indicate a ohm.m (Soltau and Anderson, 2006). paleo-weathering surface, as it is fairly level and For transect RT1 (Fig. 4a), the resistivity cross-section homogenous. It does, however, also show deeper shows a low resistivity top layer that corresponds to weathering that could be related to multiple fracture the weathered shale. The interface between the clay zones (feature 2 in Fig. 4a). North South

1 BH1 2 Weathered shale (clay)

Fractured shale grading into unweathered shale

a) North South BH3 6? 3? 4? 5?

Weathered shale (clay)

Fractured shale grading into unweathered shale

b) Southwest Northeast

7 Weathered shale (clay) 8

Fractured shale grading into unweathered shale

c)

67

www.seipub.org/ijast International Journal of Agricultural Science and Technology (IJAST) Volume 1 Issue 4, November 2013

Southwest Northeast 9? BH8 10 11?

Weathered shale (clay)

Fractured shale grading into unweathered shale

d) FIG. 4 INVERTED RESISTIVITY DATA FOR TRANSECTS RT1 (a), RT2 (b), RT3 (c) AND RT4 (d) WITH BOREHOLE POSITIONS (BLUE ARROWS) AND INTERPRETED FRACTURE FEATURES (GRAY DOTTED LINES) The resistivity cross-section for transect RT2 (Fig. 4b) drainage). Feature 10 near borehole 8 coincides with a shows a shallow subsurface layer with low to very low surface drainage pathway. resistivity that corresponds to the weathered shale. The very low resistivity that occurs in the shallow Borehole Drilling and Sampling subsurface, towards the northern end of the profile, is The geology of the catchment consists of a surficial indicative of high salinity, clay soils. The very low drape of colluvial/alluvial material overlying rocks of resistivity zones along this profile correspond to zones the Formation, which forms part of the of fracturing/preferential weathering, but this is a lot Namibian Malmesbury Group. The shales themselves less pronounced compared to transect RT1. typically vary from highly-to-slightly weathered in Undulations along the weathered surface on this character, with the effects of weathering becoming profile are less pronounced. The general depth of less pronounced with depth. The weathering weathering is also not as deep as that in the previous characteristics of the material vary from highly profile of RT1. A number of potential fracture features leached kaolinitic clay underlying the eroded silcrete have been delineated, but it was also not clear whether of the African Surface, to mottled clay in the lower these correspond to surface drainage. BH 3 coincides banks of the catchment. Unconsolidated alluvium with the most obvious fracture feature along this overlies this weathered material in the drainage profile and seems to coincide with a shallow surface feature that bisects the catchment, its depth increasing drainage channel. towards North-East. Two aquifers occur within the The inverted resistivity profile for transect RT3 (Fig. 4c) study area itself; a shallower, perched aquifer shows the shallow subsurface layer with low to very restricted to the weathered zone of the Moorreesburg low resistivity (clay) and the deeper higher resistivity Formation, and a deeper aquifer in the relatively layer (fractured shale). There is no borehole unweathered shale. Four boreholes were drilled to information to verify the interpretation along the depths between 20 and 30m, namely 1a, 5, 8a and 9a profile. Similar to RT2, the depth of weathering seems (Fig. 1c), the remaining five to depths between 40 and to be shallower. Also, the very low shallow resistivity 80m. Borehole 2 was already established on the farm. zone occurs across most of the profile and the saline The regolith salinity profiles for five boreholes are soil is more evenly developed along this profile. Two shown in Fig. 5. The data represent values of electrical potential fracture features have been indicated on the conductivity measured on saturated paste extracts profile (Fig. 4c). Feature 7 coincides with a surface (ECe) of material collected at 1 m depth increments drainage feature and this is likely controlled by the during drilling. Although boreholes 8 and 9 have subsurface fracturing. incomplete salinity profiles (Fig. 5), they are not The resistivity cross-section for transect RT4 (Fig. 4d) inconsistent with the full pattern shown for boreholes shows the shallow subsurface layer with low to very 3, 4 and 5. This pattern can be summarised as low resistivity (clay) and the deeper higher resistivity reflecting a progressive increase in salinity from low to layer (fractured shale). The low resistivity values along moderate values at the surface to a peak value at a the southwestern part of the profile coincide with the depth of between 5 and 10 m. Thereafter, salinity saline scald that has been indentified in the field and decreases asymptotically to a stable ECe value of less verified on the shallow electromagnetic survey. Two than 1 dS m-1. It was hypothesised that this decrease potential fracture features were delineated (9 and 11 in results from the lateral leaching of accumulated salts Fig. 4d; it is not clear if they correspond to surface by moving groundwater. Accordingly, salt storage

68 International Journal of Agricultural Science and Technology (IJAST) Volume 1 Issue 4, November 2013 www.seipub.org/ijast within the regolith is revealed as occurring which is in the vadose zone. A mean value of about predominantly within the vadose zone, but beneath 376 t ha-1 is therefore suggested for the small the soil zone (from which it is leached downward by catchment. In comparative terms, Brouwer and van de infiltrating rainwater) and above the saturated zone as Graaff (1988) estimated 500 t ha-1 are stored in commonly defined hydrologically. weathered regolith (pallid zone) at Gatum in Western Victoria (Australia) (in Brouwer and Fitzpatrick, 2002a and b).

FIG. 6 HYPOTHETICAL WETTING EFFECTS OF RAINWATER AND GROUNDWATER ON THE PRODUCTION OF THE SALT BULGE IN THE REGOLITH AT BOREHOLE 5, GODETROU CATCHMENT FIG. 5 REGOLITH SALINITY PROFILES FROM SATURATED PASTE EXTRACTS (ECe) OF SAMPLES OBTAINED AT 1 m DEPTH INCREMENTS FROM FIVE BOREHOLES DRILLED IN THE GOEDERTROU CATCHMENT Thus, for borehole 5, the salinity peak at about 5 m depth is illustrated in Fig. 6 due to the combined reduction in relative frequency with which the downward moving wetting front and upward migrating water table enter the zone of maximum salt accumulation. The shape and magnitude of the salinity bulge will likely depend also on the extent to which upward and downward dissolution fronts overlap, as well as the rate at, and extent to which both perched- and groundwater discharge laterally.

For reference purposes, an ECe value of 1 dS m-1 represents a soluble salt content of about 6 t ha-1 m-1. This has important implications for estimating salt storage in whole catchments. From the data for boreholes 3, 4 and 5 (Fig. 5), the regolith was calculated to contain 333, 608 and 186 t ha-1 of soluble salts for the three boreholes respectively, the bulk of

69 www.seipub.org/ijast International Journal of Agricultural Science and Technology (IJAST) Volume 1 Issue 4, November 2013

FIG. 7 TOPOGRAPHIC CONTOURS (mamsl, TOP) AND FIG. 8 TOTAL DISSOLVED SALTS (TOP) AND CHLORIDE GROUNDWATER ELEVATION INTERPOLATED FROM CONCENTRATION (BOTTOM) INTERPOLATED FROM WATER MEASUREMENTS IN THE BOREHOLES (RIGHT) IN THE QUALITY MEASUREMENTS IN THE BOREHOLES IN THE GOEDERTROU CATCHMENT GOEDERTROU CATCHMENT The rest water level and the groundwater quality were measured in the new boreholes. The interpolated groundwater level surface indicates groundwater flow from the South-West to the North-East (Fig. 7). There was a good correlation between water levels and topography with a R2 = 0.89, implying that the aquifers are semi-confined. The interpolated data map of the groundwater total dissolved salts (TDS) distribution shows a general increase in TDS with decrease in water level elevation (Fig. 8). A similar trend of increase in chloride concentration with decrease in water level elevation was also observed (Fig. 8). Significantly, groundwater quality became increasingly saline towards the bottom end of the catchment (Fig. 8). Based on chemical analyses, EC in the groundwater varied between 0.68 (borehole 1) to 3.75 dS m-1 (borehole 9). In borehole 3, EC was 1.26 dS m-1, which, in terms of the volumetric salt content it represents, is roughly consistent with ECe values of about 0.5 dS m-1 in disturbed auger samples below a depth of about 15 m (Fig. 5). Site groundwater is perhaps best classified as being of mixed cation-chloride-type, and appears to have

70 International Journal of Agricultural Science and Technology (IJAST) Volume 1 Issue 4, November 2013 www.seipub.org/ijast similar characteristics across the study area (Fig. 9). thin A horizon lies directly on shale, coincides with The piper diagram in Fig. 9 indicates that Na+ and convex shapes of the hillslope. Areas in the shallow Mg2+ dominate amongst cations, and Cl- is by far the subsurface with very low resistivity correspond to most dominant anion. Day and King (1995) indicated saline soils with high clay content. This is typical of the dominance of Na+ and Cl- in the Berg river weathered shale. Significantly, weathered areas of the catchment. For extensive data on the discharge of salts profile correspond to subsurface geological and, in from the Goedertrou catchment via overland flow and most cases, topographic flow pathways. interflow, the reader is referred to Jovanovic, Bugan, The electromagnetic induction data indicated lower Frantz, De Clercq and Fey (2008) and De Clercq, soil salinity at the end of the rainy winter season Jovanovic and Fey (2010). compared to the dry summer season. The impact of salt discharge was therefore lessened by the fact that most of it was released during periods of high flow so that considerable dilution occurred. The salinity hazard was, however, probably magnified seasonally in the Berg river basin by evaporation from the river and storage dams and by more concentrated salt seepage into the lower reaches during periods of low flow and greater water demand. The factors related to the magnitude and shape of the deep salinity bulge in the vadose zone need to be investigated further since it may hold the key to estimating the quantity of stored salts in the drier regolith of the lower Berg river basin. Different FIG. 9 PIPER DIAGRAM OF GROUNDWATER CHEMISTRY AT climates, hydrogeological characteristics and land uses GOEDERTROU may have resulted in different amounts of salt storage in the soils and regolith. Both land use responses and Conclusions salt storage capacity are key factors for hydrological This study allowed us to converge on the modelling and catchment management. It will be characterization of regolith salt storage in a small essential to investigate a variety of land uses and catchment of the Berg river basin through intensive management in order to recommend the most electromagnetic induction and geophysical favourable ones in terms of water consumption and measurements, as well as borehole information. The salt discharge. regolith in this semi-arid coastal region contains an abundance of stored salts. The regolith in three Groundwater quality becomes increasingly saline boreholes showed a sharply peaking concentration of (mixed cation-chloride-type) from the recharge area soluble salts between about 5 and 10 m depth, due to towards the bottom end of the catchment. This rate of the wetting and dissolution combined effects of salt discharge is sensitive to changes in both climate infiltrating water and rising water table. As a rough and agricultural practice and the current rate has the estimate, the regolith contains between 186 and 608 t potential to persist, to the detriment of water users, for of salts ha-1, stored mainly as a bulge in the decades at least, if not a century or two. It would unsaturated vadose zone above the water table and therefore make sense to explore ways in which below the soil zone, discharging mainly through reduced discharge (i.e. enhanced storage) can lateral leaching by groundwater from the phreatic economically be engineered in key parts of the basin. zone. ACKNOWLEDGMENT Flow pathways and mechanisms of salt transport in the small catchment were identified through The authors acknowledge the Water Research electromagnetic induction and resistivity measurements. Commission (Pretoria, South Africa) for funding the Salinity in the landscape is driven by the topo- research; the Western Cape Office of the Department sequence with anti-erosion contours acting as barriers of Agriculture (Elsenburg, South Africa) for funding to water and salt fluxes. Presence of highly saline the borehole drilling; JJ Myburgh and Brothers for scalds, where the soil B horizon is absent and a very drilling the boreholes.

71 www.seipub.org/ijast International Journal of Agricultural Science and Technology (IJAST) Volume 1 Issue 4, November 2013

REFERENCES rivers.” South African Journal of Science 91 (1995): 299-

Acworth, R.I. “Investigation of dryland salinity using the 306. electrical image method.” Australian Journal of Soil De Clercq, W.P., Jovanovic, N., Fey, M.V. “Land use impacts Research 37 (1999): 623-636. on salinity in Western Cape waters.” WRC report No. Acworth, R.I., Jankowski, J. “Salt source for dryland salinity 1503/1/10. Pretoria: Water Research Commission, 2010. – evidence from an upland catchment on the Southern DWAF. “Hydrology of the Berg River Basin.” Department of Tablelands of New South Wales.” Australian Journal of Water Affairs and Forestry report No. PG000/00/2491. Soil Research 39 (2001): 39-59. Pretoria: R.R. Berg (Ninham Shand) in association with Angus, J.F., Gault, R.R., Peoples, M.B., Stapper, M., Van BKS Inc., 1993. Herwaarden, A.F. “Soil water extraction by dryland DWAF. “National Water Resource Strategy. 1st edn.” Available crops, annual pastures, and lucerne in south-eastern at http://www.dwaf.gov.za/Documents/Policies/NWRS/ Australia.” Australian Journal of Agricultural Research Default.htm [accessed 25 March 2009], 2004. 52 (2001): 183–192. Engel, R., McFarlane, D.J., Street, G. “Using geophysics to Bennett, D.L., George, R.J. “Using the EM38 to measure the define recharge and discharge areas associated with effect of soil salinity on Eucalyptus globules in south- saline seeps in south-western Australia.” In Groundwater western Australia.” Agricultural Water Management 27 Recharge, edited by M. Sharma, 20-40. Rotterdam: (1995): 69-86. Balkema Publishing Company, 1989. Brouwer, J., Fitzpatrick, R.W. “Interpretation of FAO. “World Reference Base for soil resources.” Rome: Food morphological features in a salt-affected duplex soil and Agricultural Organization of the United Nations, toposequence with an altered soil water regime in 1998. western Victoria.” Australian Journal of Soil Research 40 Fey, M.V., De Clercq, W.P. “Dryland salinity impacts on (2002a): 903-926. Western Cape rivers.” WRC report No. 1342/1/04. Brouwer, J., Fitzpatrick, R.W. “Restricting layers, flow paths, Pretoria: Water Research Commission, 2004. and correlation between duration of soil saturation and Flugel, W.A. “River salination due to dryland agriculture in soil morphological features along a hillslope with an the Western Cape Province, Republic of South Africa.” altered soil water regime in western Victoria.” Australian Environmental International 21(5) (1995): 679-686. Journal of Soil Research 40 (2002b): 927-946. Gorgens, A.H.M., De Clercq, W.P. “Summary of water Brouwer, J., Van De Graaff, R.H.M. “Readjusting the water quality information system and soil quality studies.” balance to combat dryland salting in southern Australia: WRC report No. TT252/05. Pretoria: Water Research changing the hydrology of a texture contrast soil by deep Commission, 2005. ripping.” Agricultural Water Management 14 (1988): 287- Greiner, R. “Catchment management for dryland salinity 298. control: Model analysis for the Liverpool Plains in New Bussian, A.E. “Electrical conductance in a porous medium.” South Wales.” Agricultural Systems 56(2) (1998): 225-251. Geophysics 48(9) (1983): 1258 – 1268. Hingston, F.J., Gailitis, V. “The geographic variation of salt Clarke, C.J., George, R.J., Bell, R.W., Hatton, T.J. “Dryland precipitated over Western Australia.” Australian Journal salinity in south-western Australia: its origins, remedies, of Soil Research 14 (1976): 319–335. and future research directions.” Australian Journal of Hughes, J.D., Khan, S., Crosbie, R.S., Helliwell, S., Michalk, Soil Research 40 (2002): 93-113. D.L. “Runoff and solute mobilization processes in a Corwin, D.L., Plant, R.E. “Applications of apparent soil semiarid headwater catchment.” Water Resources electrical conductivity in precision agriculture.” Research 43 (2007): 1-15 (W09402). Computers and Electronics in Agriculture 46 (2005): 1-10. Johnson, C.D. “Distribution of environmental chloride in Day, J.A., King, J.M. “Geographical patterns, and their relation to subsurface hydrology.” Cited in Peck, A.J., origins, in the dominance of major ions in South African Williamson, D.R. “Hydrology and salinity in the Collie

72 International Journal of Agricultural Science and Technology (IJAST) Volume 1 Issue 4, November 2013 www.seipub.org/ijast

River Basin Western Australia.” Journal of Hydrology 94 Australia.” The Australian Journal of Agricultural and (1987): 67–88. Resource Economics 45(3) (2001): 459-475. Jovanovic, N.Z., Bugan, R.D.H., Frantz, G., De Clercq, W.P., Parsons, R.P. “A South African aquifer system management Fey, M.V. “Hydrosalinity fluxes in a small scale classification.” WRC report No. 77/95. Pretoria: Water catchment of the Berg river (South Africa).” In Water Research Commission, 1995. Pollution IX edited by D. Prats Rico, C.A. Brebbia, Y. Soil Classification Working Group. “Soil classification. A Villacampa Esteve Y, 603-612. Southampton: Wit Press, taxonomic system for South Africa.” Pretoria: Department 2008. of Agricultural Development, 1991. Loke, H. “Rapid 2D and 3D resistivity and IP inversion Soltau, L.E., Anderson, N.L. “An integrated geophysical using the least squares method. RES2DINV v. 3.4 survey of groundwater resources, Mamre area, Western manual.” Available at www.geoelectrical.com [accessed Cape Province.” South African Journal of Geology 109(4) 23 March 2009], 2001. (2006): 433 – 438. McArthur, W.M., Churchward, H.M., Hick, P.T. Stirzaker, R.J., Cook, F.J., Knight, J.H. “Where to plant trees “Landforms and soils of the Murray River catchment on cropping land for control of dryland salinity: some area of Western Australia.” Land resources and approximate solutions.” Agricultural Water Management management series No. 3. Canberra: CSIRO, 1977. 39 (1999): 115-133. McNeill, J.D. “Electrical conductivity of soil and rocks.” Telford, W.M., Geldart, L.P., Sheriff, R.E. “Applied Technical note TN-5, Geonics Ltd., Missisauga, Ontario, geophysics.” 2nd edn. Cambridge: Press Syndicate of the Canada, Available at http://www.geonics.com/pdfs/ University of Cambridge, 1990. technicalnotes/tn5.pdf [accessed 26 March 2009], 1980. Van Rensburg, L.D., De Clercq, W.P., Barnard, J.H., Du Pannell, D.J., Ewing, M.A. “Managing secondary dryland Preez, C.C. “Salinity guidelines for irrigation: Case salinity: Options and challenges.” Agricultural Water studies from Water Research Commission projects along Management 80 (2006): 41–56. the Lower Vaal, Riet, Berg and Breede Rivers.” Water SA Pannell, D.J., McFarlane, D.J., Ferdowsian, R. “Rethinking 37(5) (2011): 739-750. the externality issue for dryland salinity in Western

73